914 



SCIENCE 



[N. S. Vol. XLII. No. 1095 



of cancer (1) hereditary factors, and (2) 

 chemical actions exerted by an internal secre- 

 tion. 



We coTild show that the hereditary factors 

 are not identical with the internal secretion 

 and do not act by changing the number of 

 corpora lutea and their activity, but that their 

 point of attack is somewhere else. It appears 

 probable that with the cooperation of heredi- 

 tary conditions all those internal secretions are 

 factors in the origia of cancer which initiate 

 or sustain continuous or periodic growth 

 processes. In other cases mechanical stimu- 

 lation of growth may take the place of chem- 

 ical stimulation and again in others a com- 

 bination of both may be present. Whether ia 

 addition to these factors definitely established 

 there is still another factor (microorganisms ?) 

 present, and which relation this last-named 

 hypothetical factor bears to the other two 

 factors are at present unknown. But whether 

 or not such an additional factor enters, we can 

 be certain that the two first named sets of fac- 

 tors are sufficiently strong to determine to a 

 great extent the frequency of cancer in mice. 

 Leo Loeb 



Depaetment op Compakative Pathology, 

 Washington TJniveksitt 



a new method of selecting tomatoes foe re- 

 sistance to the wilt disease 

 Perhaps the most serious disease of toma- 

 toes in the southern states is that caused by 

 Fusariutn lycopersici, the one that is commonly 

 known as the tomato wilt. The causative 

 fungus lives in the soil and attacks the plants 

 through the roots and later grows up through 

 the fibrovascular bundles into the stems. In 

 common with similar diseases of other plants, 

 such as the wilts of cotton, watermelon, etc., the 

 only practical method of control now known 

 is in the use of varieties, or strains, that are 

 resistant to the disease. By saving seed from 

 healthy plants in a badly infected field for sev- 

 eral seasons, strains can be obtained which 

 show considerable resistance to the disease. 

 This method of selecting the strains, however, 

 has several serious drawbacks: (1) A large 



acreage of tomatoes is required as a large per- 

 centage of the plants which are set in the field 

 die with the wilt. (2) Many of the plants in 

 the field do not come in contact with the wilt 

 fungus during the season and so do not have 

 a chance to show whether they are resistant to 

 the disease. (3) Resistant plants in the field 

 are readily pollinated by the susceptible 

 plants. (4) The time necessary to obtain a 

 wilt-resistant strain is too long. 



While studying the disease in Louisiana, an 

 attempt has been made to improve on our com- 

 mon method of selecting resistant plants by 

 selecting in the seed bed. This has been ac- 

 complished by taking advantage of the fact 

 that soil diseases infect plants better if the 

 soil is first sterilized and then inoculated with 

 a pure culture of the disease organism. In 

 ordinary unsterilized garden soil, even if it is 

 heavily inoculated with the tomato wilt 

 fungus, not many of the plants will show the 

 wilt to any extent before it is time to place 

 them in the field. The presence of bacteria 

 and other fungi seems to have an inhibitory 

 effect on the wilt fungus. If, however, the soil 

 is first sterilized by heat and then heavily in- 

 oculated with the wilt fungus just before 

 planting, the disease wUl develop so well that 

 most of the susceptible plants will be killed 

 before they are large enough to be placed in 

 the field. By growing the plants in this man- 

 ner, only plants showing resistance are placed 

 in the field. This saves a great deal of field 

 space and also allows a selection from an al- 

 most unlimited number of plants. This also 

 assures the presence of the wilt fungus on the 

 roots of every plant. 



To show how this method of selection works 

 in practise, results of some experiments may 

 be briefly given. Having by the old method of 

 selection obtained a strain that showed con- 

 siderable resistance to the wUt disease, this 

 was compared by the seed bed method with 

 three standard varieties of tomatoes. The seed 

 of each variety were planted side by side in 

 reinoculated sterilized soil. Different cultures 

 of the fungus from different localities were 

 also used in order to see if they would affect 

 the varieties differently. In the following 



